JPH0974219A - Surface light-emission type diode and its manufacture - Google Patents

Abstract

PROBLEM TO BE SOLVED: To increase the high brightness coupling efficiency of an LED by a method wherein a block layer, which is selectively oxidized, is formed as far as to the mesa side face of the second conductive type thin film excluding the center part. SOLUTION: After a mesa etching treatment has been conducted on the part as far as to a clad layer 5 (2), an oxide is selectively grown on a part of the part 6 (2) of an AlGaAs layer. At this time, the growth of an oxide from mesa edge has no relation with time, and it is selected so as to have an aperture part (non-oxide part) 6 (1) which becomes equal to the diameter of a light emitting surface 12. A current is applied to a junction via the aperture part 6 (1). For example, when the thin film of the AlGaAs layer 6 is used, a III-V compound semiconductor, on which said thin film is lattice-matched with a clad layer and a selective oxide can be grown (AlAs for example), is used. After an etching treatment has been conducted, an insulating layer 9 is deposited. Subsequently, a window is perforated on the insulating layer 9 for the light emitting surface. Then, p-type contact electrodes 10 (1) and 10 (2) are formed.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an optical device used in an optical communication system, and more particularly to an optical device
The present invention relates to a high-brightness visible LED used as a light source with high coupling efficiency in a POF based on a data link system. This LED can also be used as an outdoor display device or an indicator of an automobile.

[0002]

The light output efficiency of conventional light emitting diodes (LEDs) is related to the structure and shape of the electrodes used to inject current into the junction. All surface-emitting LEDs that have been manufactured thus far have a rectangular or circular upper electrode provided in the center. The area and shape play an important role in the light output efficiency, and in order to increase the light output efficiency, it is required that the electrode area is small and the light emitting surface is wide.

In conventional LEDs, the use of a centrally located rectangular / circular electrode facilitates manufacturing, but it splits the Gaussian pattern of the output beam, especially diverging the beam to create a radiation pattern at the electrode periphery. Concentrate. The radiation pattern decreases as the distance from the electrode increases. This is because the current is concentrated in the central part and is not distributed far from the electrodes. The beam intensity achieved is directly proportional to the current density at the location. Conventional LEDs with such a wide beam pattern cannot be used in optical data link systems. The reason is that even if a large core fiber is used, the coupling efficiency becomes low due to its radiation pattern.

A typical example is shown in FIGS. 16 and 17. 16 (a) and 16 (b) show the top surface of the LED and its cross-section along the approximate near-field radiation. In FIG. 16, the n-type GaAs buffer layer 2, the n-type (Al _x G
a _1-x ) In _1-x P layer 3, active layer 4, p-type (Al _x Ga
_1-x ) _0.5 In _0.5 P layer 5, current injection layer 7, p-type GaAs
The s cap layer 8 is continuously grown on the n-type GaAs substrate 1. Conventionally, a rectangular, circular, or cross-shaped electrode 28 is used for the p-type contact. As the current injection layer 7, thick Al _0.7 Ga _0.3 As is usually used.

FIG. 17 shows a schematic near-field pattern of intensity 30 for an LED having a centrally located rectangular or circular electrode. Light intensity 3 as shown
0 is maximum around the edge of the electrode and is low at a position away from the electrode. This indicates that the current is mainly concentrated under the electrode contact. In this case, it is not possible to boost the light as expected. In order to avoid current concentration around the electrodes, a block layer is usually used before the AlGaAs layer (current injection layer) 7. FIG.
Shows another example of a conventional LED having a block layer 31. The need for regrowth increases the manufacturing cost of LEDs. In addition, the darkness of the center for circular electrodes reduces coupling efficiency even with large core fibers. It is highly desirable to provide LED structures that provide high coupling efficiency and are suitable for mass production.

A visible LED (580-67) made of a III-V semiconductor having a rectangular / circular electrode provided in the center.
A wavelength of 0 nm) is disclosed in many documents and patent publications. In either case, the rectangular or circular electrode provided in the center is used as the upper contact for current injection. One of the representative examples is the paper “Japan Journal of Ap” by Sugawara et al.
plied Physics, Part 1, Vol. 3
1, no. 8, pp-2446-2451, 1992 "
Can be found at In this paper, cross-shaped electrodes for current injection are used in surface emitting LEDs. In addition, a blocking layer is used to prevent current from concentrating under the contacts.

However, the light output efficiency has not yet exceeded its practical application. A thick window layer with low resistance is required to increase current injection out of the contact. Even with this thick layer, the current injection out of the contact is limited to a certain level and the light intensity is quite low for practical applications. This kind of LED is
It may be fine for use in outdoor applications. However, the use of conventional LEDs exhibits low coupling efficiency, especially in applications such as short range data link systems based on POF, and conventional LE for communication system based POF.
D is not practical. PO based on application of data link
For F, it is highly desirable to design an LED that not only has high brightness but also exhibits high coupling efficiency. Furthermore, the manufacturing cost must be as low as possible to reduce the cost of the system.

Japanese Unexamined Patent Publication No. 2-174272 discloses a high brightness LED. According to the present invention, an n-p-n-p structure having a structure that helps spread the current out of the contact
The structure is adopted to increase the brightness of the LED. The main drawback of this invention is that a mesa structure up to the active region must be created in the light emitting surface to create a current path before the p-type contact is made, and a dopant such as zinc (Zn) is required. Is that it must diffuse in a high temperature environment. Z
Post-growth high temperature treatment for n-diffusion affects its performance characteristics due to degradation of the active region. Since a highly p-doped emitting surface is used, most of the light (due to the energy of the light) may be absorbed by the diffusion layer. LE of this kind
The manufacture of D requires several treatments, which increases the cost of manufacturing the LED. Not only that, it is also difficult to make high speed LEDs due to the high capacitance generated due to the large contact area.

Japanese Unexamined Patent Publication No. 3-190287 discloses a high brightness LED array. In this LED array, contacts are made in the top layer following mesa formation to facilitate wire bonding. In this case, the shape of the contact has been modified to achieve high light output. The present invention relates to an LED array and requires a mesa structure before forming electrodes, so that it is difficult to form electrodes unless a thick contact is formed. Not only is the manufacturing cost of the LED increased, but it is not entirely practical to make a single LED.

Japanese Unexamined Patent Publication No. 5-211345 discloses LE
Discloses D. In the present invention, an LED is manufactured using a block layer of the opposite conductivity type (p-type or n-type) of a compound semiconductor and diffusion at high temperature. Furthermore, this L
The top electrode geometry used in EDs is impractical for use as a light source in optical data link systems due to low coupling efficiency. Furthermore, high temperature treatments degrade optical properties due to doping diffusion and crystal defects.

Japanese Unexamined Patent Publication No. 4-174567 discloses a high brightness LED array used in a printer. In this invention, a distributed Bragg reflector layer (DBR) at the bottom is used to reflect the return light from the substrate. In this case, the same idea as manufacturing a surface emitting laser is implemented. For the DBR, a GaAs / AlAs pair is used to reflect light with a wavelength of 880 nm and go to the substrate. The type and number of pairs used in the DBR must be independent of the output radiation wavelength and this DBR
Cannot be used for visible LEDs involving wavelengths of 600-650 nm. This LED array can be used in printers, but not in data link systems. This is because the shape of the upper electrode is the same as that of the conventional LED described above, and the coupling efficiency with the fiber is extremely low as a result of using this LED.

Japanese Unexamined Patent Publication No. 4-259263 discloses In
A visible LED using AlGaP is disclosed. In the present invention, an additional layer of InGaP having a bandgap wider than that of the layer used for the active layer is used to avoid Zn diffusion into the active region. No additional layer for current injection is used. As mentioned above, this type of visible LED with a circular electrode in the center is
It cannot be used for short range data link systems.

As mentioned above, the previously proposed L
The optical output and coupling efficiency of ED structures are insufficient for use in short range data link systems, especially where POF based systems are concerned. Therefore, it is highly desired to design a high-luminance visible LED that can be manufactured at low cost and is suitable for mass production.

[0014]

[Problems to be Solved by the Invention]
Must be mitigated for potential applications in F, or silica based data link systems, or other display systems.

These problems can be mitigated by evenly distributing the current on the light emitting surface. If the current is prevented from spreading toward the bonding region, the current can be injected only into the junction of the light emitting surface. A circular emitting surface is used to match the shape of the output beam with the fiber core, thereby increasing coupling efficiency. Optimize the diameter of the emitting surface (the choice depends on the fiber dimensions)
As a result, a uniform light emission pattern can be realized. The LEDs of the present invention are also useful in display systems.

[0016]

BEST MODE FOR CARRYING OUT THE INVENTION Embodiments of the present invention will be described in detail with reference to the drawings.

[0017]

[Embodiment 1] FIG. 1 is an explanatory view showing the structure of a visible LED according to the first embodiment. (A) is sectional drawing, (b) is a top view. As shown, the buffer layer n-GaA
s layer 2, n- (Al _0.7 Ga _0.3 ) _0.5 In _0.5 P layer 3, active layer 4, p- (Al _0.7 Ga _0.3 ) _0.5 In
_0.5 P layer 5 and Al _x Ga _1-x As layer 6 (x ≧ 0.9)
Thin layer, a current injection layer Al _0.7 Ga _0.3 As layer 7, and a highly doped p-GaAs cap layer 8 on the substrate n-GaA.
The s layer 1 is continuously grown. As the active layer 4, using a bulk or multi-quantum well _{(Al x Ga 1-x)} 0.5 In 0.5 P was based material having a desired wavelength. Al _x Ga _1-x As (x ≧ 0.7) is used as the current injection layer 7. The current injection layer is a p-type III having a bandgap wider than that of the material used as the active layer.
It is a -V compound semiconductor and lattice-matches with the lower p-type cladding layer.

During the epitaxial growth, the current injection layer 7
Since the cap layer 8 needs to be highly doped, the diffusion of the dopant into the active layer deteriorates the optical characteristics of the LED. In order to prevent the dopant from diffusing into the active layer, a spacer layer 5 (1) having a thickness based on the thicknesses of the current injection layer 7 and the cap layer 8 is formed subsequent to the active layer 4. As the spacer layer, the same low-doped p-type material used as the p-type cladding layer 6 can be used.

After the mesa etching up to the clad layer 5 (2), selective oxide growth is performed on a part of the AlGaAs layer 6 (2). This can be done by heating the sample at a temperature of 400 ° C. or higher in a steam environment realized by water bubbling. AlGaAs A
Since the composition ratio of 1 is larger than 0.9, selective oxide growth can be performed on the layer 6 (2). Oxide growth from the mesa edge is time independent and must be chosen so that the opening (non-oxide portion) 6 (1) is equal to the diameter of the emitting surface 12.

Current is injected into the junction through opening 6 (1). In the preferred embodiment, a thin layer of AlGaAs layer 6 is used, which is lattice matched to the cladding layer 5 and is capable of selective oxides (eg AlAs) III-
A group V compound semiconductor is used. After epitaxial growth, 1
Silicon nitride or silicon oxide 9 with a thickness of 50 nm or more is deposited at a temperature of 500 ° C. or less. Subsequently, a window for the light emitting surface 12 is opened in the insulator layer 9. Next, the p-type contact electrodes 10 (10 (1) and 10 (2)) are formed using the lift-off method.

In LEDs, the current density is higher than that of optical devices such as laser diodes, so p
The contact resistance mainly caused by the shaped electrodes must be as low as 1 ohm for long-term reliable operation. For this purpose, as a p-type contact, G
More reliable metallization must be chosen to avoid the formation of any spikes in the aAs cap layer. Au: Zn or Ti / Pt / Au or Ni / Z as p-type contact
n / Au can be used and exhibits considerably higher reliability compared to the usual types of metallizations such as Cr: Au.

To use LEDs in a data link system, the optical coupling efficiency with the fiber is considered. For this reason, the emitting surface must be designed so that maximum coupling efficiency is achieved. Coupling efficiency is mainly related to several factors: the type of fiber used and the emitting surface of the LED used. In the case of LEDs, the emitting surface must be designed so that more light is concentrated in the central area than in the outside. And, the coupling efficiency is more related to the choice of the diameter of the emitting surface (ie the diameter of the upper electrode). The coupling efficiency can be maximized by optimizing the diameter of the top electrode corresponding to the light emitting surface.

As described above, when the conventional electrode provided mainly in the central portion is used, the output light is expanded to the outer edge of the LED, which lowers the coupling efficiency. If a ring-shaped electrode and a current blocking below the bonding area 16 are used, the light can be concentrated in the center, which results in a higher coupling efficiency than with conventional LEDs. Can be higher. The design of the circular electrode along with the circular light emitting surface is also an important factor to increase the coupling efficiency and its brightness. In order to increase the output power, the electrode diameter (the diameter of the light emitting surface) must be optimized.

FIG. 2 shows how the diameter of the light emitting surface relates to the light output efficiency as a function of the thickness of the current injection amount (AlGaAs). The use of a thick current injection layer gives higher brightness for a given light emitting surface area. But,
With a wider emitting surface, the light output efficiency approaches that of an LED with a thick current injection layer. To achieve maximum coupling efficiency (≧ 50%) with the fiber, the diameter of the emitting surface should be designed such that the ratio of the diameter of the fiber core to the emitting surface is 5 or more. If a step index (SI) POF fiber with a core size of 0.98 mm and a numerical aperture (NA) of 0.5 is used, the ring diameter (emission surface) used will give a coupling efficiency of 80% or more. To achieve it, it must be as small as 100 μm or less.

After making the p-type contact, the wire
Au10 (2) was added to facilitate bonding.
Plate on the p-contact or on the part where wire bonding will be done. Subsequently, the light emitting surface 12 is covered with the passivation layer 13. Alumina (Al ₂ O ₃ ), silicon oxide (SiO ₂ ), or silicon nitride (SiN _x ) can be used for the passivation layer 13. The passivation layer is at room temperature or 20
It must be deposited at temperatures below 0 ° C. This passivation layer 13 prevents the light emitting surface 12 from absorbing water molecules or oxygen from the surroundings, giving the device reliability for long term operation. After forming the passivation layer, the back surface of the substrate is polished by about 150 μm to produce the n-type contact 11. The n-type contact 11 has A
u: Ge / Ni / Au can be used. After making the n-type contact on the back side of the polished substrate,
To scribe for packaging. The use of a blocking layer (in this case selective AlGaAs oxide) and the simple structure of the LED not only improve the light output power and coupling efficiency, but also minimize the manufacturing cost of the LED.

FIG. 3 is a diagram showing an LED of the second embodiment. In FIG. 3, the same reference numerals as in the first embodiment are used. In the second embodiment, a distributed Bragg reflection layer (DBR) mirror 14 is grown on the buffer layer 2 before growing the cladding layer 3, active layer 4 and current injection layer 7. The DBR used as a mirror is composed of a pair of semiconductors 14 (1) and 14 (2) having a high reflectance difference. The number and type of semiconductor pairs determines the level of reflection of the DBR mirror. As the semiconductor layer used as the DBR, one having a low absorption property for radiated light is used. FIGS. 4A and 4B show an explanatory example showing the DBR together with its schematic reflectance characteristics. As described above, the number of semiconductor pairs and the types used for the semiconductor pairs are related to the wavelength of output light. In the case of optical devices, especially LEDs, the emission spectrum covers a wide wavelength range.
R must have a high reflectance in a wide wavelength range. As shown in FIG. 4 (b), maximum reflectance must be achieved within a wide wavelength range of λ ₁ to λ ₂ . This helps to achieve maximum reflectivity over the wafer area, while D due to thickness non-uniformity (if any).
There is a BR reflectance shift. When designing a 650 nm LED, AlAs, Al _x G
a _1-x As (x ≧ 0.45), (Al _x Ga _1-x ) _0.5
In _0.5 P (x ≧ 0.45), Ga _0.5 In _0.5 P, or GaAs can be used. A wide wavelength range can be realized by using a single DBR or an individual DBR.

Tables 1 and 2 are a summary of the semiconductor pair types and DBR characteristics that can be used when designing a visible LED with a wavelength of 650 nm.

[0028]

[Table 1]

[0029]

[Table 2]

When designing a DBR, it is important to form a graded junction to minimize its resistance, as well as its reflective properties. To this end, superlattices or graded junctions that match the semiconductor pairs are used. Following the formation of the DBR 14, as shown in FIG. 3, the n-type cladding layer 3, the active layer 4,
The p-type cladding layer 5, the current injection layer 7, and the p-GaAs cap layer 8 are continuously grown in one chamber. Other processes subsequent to this step are the same as those described in the first embodiment, and therefore repeated description is omitted here.

FIG. 5 is a diagram showing the structure of the LED of the third embodiment. In this figure, the same members are designated by the same numbers as in the first embodiment. Therefore, it will not be described again. In the third embodiment, the mesa etch is unnecessary. Prior to depositing the insulator layer 9 such as silicon oxide or silicon nitride, the light emitting surface 12 is coated with a thick resist or other kind of material that can be selectively etched, followed by a proton implantation 15. Proton implantation is done to create a high resistance region and avoid spreading the current towards the bonding region 16. The selective oxide layer 6 (2) used in the first embodiment is not required in the third embodiment. Using ion implantation with the blocking layer helps to increase light output and coupling efficiency.
The subsequent process has already been described in the first embodiment, and will not be described again.

FIG. 6 is a diagram showing the LED structure of the fourth embodiment. In the figures, the same parts are designated by the same reference numbers in the first, second and third embodiments and are therefore not described again. In the fourth embodiment, the DBR14
After the formation of, the n-type cladding layer 3, the active layer 4, and the p-type cladding layer 5 are continuously grown. The current blocking region is created by ion implantation 15. The process following this has already been described in the second embodiment, so a repetitive description will be omitted.

FIG. 7 shows the LED structure of the fifth embodiment. In the figures, the same parts are designated by the same reference numerals in the first embodiment and are therefore not described repeatedly. In the fifth embodiment, the thin layer 6 used in the first embodiment is not used. In this embodiment, a current is injected into the junction using a mesa structure having an area slightly larger than the light emitting surface. The bonding portion 17 is outside the mesa structure. After the mesa etching, the insulating layer 9 is deposited. The process that follows is
Since it has already been described in the first embodiment, the repetitive description will be omitted. Due to its simple structure, this LED can be made at low cost and it is suitable for mesa scale manufacturing.

FIG. 8 shows the LED structure of the sixth embodiment. In the figures, the same parts are designated by the same reference numbers in the first, second and fifth embodiments and are therefore not described repeatedly. In the sixth embodiment, DBR14
Of the n-type clad layer 3, the active layer 4, the p-type clad layer 5, the current injection layer 7 and the cap layer 8,
Grow continuously on BR. The process that follows is
Since it has already been described in the first embodiment, the repetitive description will be omitted.

FIG. 9 shows the LED structure of the seventh embodiment. In the figures, the same parts are designated by the same reference numerals in the first embodiment and are therefore not described repeatedly. In this example, the thin layer 6 and the mesa etch are required. In this seventh embodiment, two epitaxial growth steps are required to make the LED.
After forming the p-type clad layer 5, the block layer 18 is grown, and then a pattern is formed. The block layer 18 is
It helps prevent current from concentrating under the bonding region 16. As the block layer 18, the clad layer 3
The same n-type material can be used. Block layer 18
After patterning, a current injection layer 7 and a heavily doped GaAs cap layer 8 for p-type contact are subsequently grown. The subsequent process has already been described in the first embodiment, and will not be described again. The use of the blocking layer helps to inject current into the junction without spreading it towards the bonding layer, increasing light output efficiency.

FIG. 10 is a diagram showing the LED structure of the eighth embodiment. In the figures, the same parts are designated by the same reference numbers in the first, second and seventh embodiments and are therefore not described repeatedly. In the eighth embodiment, DBR1
4, the n-type clad layer 3, the active layer 4, the p-type clad layer 5, and the block layer 18 are continuously grown.
The process following this has already been described in the second embodiment, so a repetitive description will be omitted.

FIG. 11 shows the LED structure of the ninth embodiment. In the figures, the same parts are designated by the same reference numerals in the first embodiment and are therefore not described repeatedly. The thin layer 6, mesa etching and passivation layer 13 used in the first embodiment are not used in this embodiment. In the sixth embodiment, a thin layer 19 of GaAs about 100 angstroms thick is used to make a p-type contact to avoid absorption of emitted light. Further, a ring (single or double or more) -shaped upper electrode is used, and ions are injected into the current injection layer to form a high resistance region, and the current is uniformly diffused in the current injection layer 7. Following the deposition of the insulating material 9, a ring-shaped window is provided in the insulating material for Au-Zn or Ni / Zn / Au for p-type contacts.
Alternatively, Cr: Au is deposited. A ring-shaped electrode is formed on the light emitting surface 20 by lift-off. Then, Cr: Au is vapor-deposited on the bonding region 21, and Au plating 22 is performed on all the electrode parts to manufacture an LED. Since an insulating material is used for the bonding portion and the passivation layer, the insulating material should be selected so that absorption of emitted light does not occur. For example, for 650 nm LEDs, silicon oxide or alumina can be used as the insulating material 19. According to this embodiment, the manufacturing cost of the chip can be minimized since only a few steps are required to manufacture the LED. In this embodiment, the high resistance region is formed in the current injection layer by ion implantation, but it may be an oxide block layer or an insulating block layer.

FIG. 12 is a diagram showing an LED of the tenth embodiment. In the figures, the same parts are designated by the same reference numbers in the first, second and ninth embodiments and are therefore not described repeatedly. In the tenth embodiment, DBR1
4, the n-type clad layer 3, the active layer 4, the p-type clad layer 5, the current injection layer 7, and the GaAs thin layer 19 are continuously grown. The subsequent process has already been described in the first and ninth embodiments, and will not be described again.

FIG. 13 is a view showing the LED mold structure of the eleventh embodiment. In the eleventh embodiment, after bonding the LEDs on the leaf frame 24, a molding 25 is performed to increase the coupling efficiency. The lens 26 is produced when the molding 25 is performed.
In order to improve the coupling efficiency and keep the packaging cost as low as possible, the lenses used are made of the same type of material as the molding material. In data link applications based on POF 27, the radius (R) of lens 26 is ≤
The distance from the center 26 (c) of the lens to the LED should be as small as 0.75 mm or less. The material used for molding has a refractive index of n ≦ 1.60 and must be made at a temperature of 200 ° C. or lower.

14 and 15 show the evaluation results of the molded lens shown in FIG. It can be seen that if the LED 23 with the mold 25 is used in the step index (SI) POF 27 having a numerical aperture (NA) of 0.5 or less and a diameter of 0.9 mm or less, the coupling efficiency can be increased to 80% or more. The position of the LED 23 with respect to the lens center 26 (c), the lens distance Z ₁ , and the fiber distance Z ₂ (X
₁ ) plays an important role in coupling efficiency. Coupling efficiency 80
In order to obtain at least 0.1%, Z ₂ is preferably 7.5 mm or less. To achieve coupling efficiency of 90% or more, the minimum lens distance Z ₁ must be 0.7 mm.

The preferred embodiment described above describes an LED that can be used in a POF-based optical system. SI POF was used to demonstrate the feasibility of using the LED 23 of the present invention as a light source for a short range data link system. However, it is also possible to use the LEDs of the invention in systems based on all types of POF or regular silica fibres. It can also be used for outdoor applications. The LED structure described in the preferred embodiment can also be applied to LEDs with a wavelength of 0.5 to 1.6 μm.

A description of the preferred embodiment of the present invention is given below.
It is provided for the purpose of illustration and is not intended to limit the invention to the disclosure. The foregoing description has been chosen and described in order to best explain the principles of the invention and may be used in various embodiments,
Various modifications may be made that are suitable for the particular intended use.

[0043]

The visible LED having the block layer below the bonding area has a higher output power than the LEDs which have been conventionally developed. The LED of the present invention is
Further, it showed a binding efficiency of 80% or more with POF. The use of a circular electrode with a circular surface helps to spread the current evenly outside the contact, making the LED brighter and increasing the coupling efficiency over conventional LEDs.

[Brief description of drawings]

FIG. 1 is a diagram showing an LED in a first embodiment.

FIG. 2 is a graph showing calculation results showing the dependence of the light emitting surface size on the light output efficiency.

FIG. 3 shows a second example LED having a DBR and a selective oxide layer as a blocking layer.

FIG. 4 is a diagram showing the structures of DBRs used in the second, fourth, sixth, eighth and twelfth examples together with their reflectance characteristics.

FIG. 5: L of the third embodiment having an ion implantation block layer
It is a figure which shows ED.

FIG. 6 is a diagram showing a fourth embodiment LED having an ion implantation block layer and a bottom DBR.

FIG. 7 L of the fifth embodiment having mesas etched to avoid current spreading underneath the bond.
It is a figure which shows ED.

FIG. 8 shows a structure of a mesa-etched sixth embodiment LED having a bottom DBR.

FIG. 9 is a view showing an LED of a seventh embodiment having a block layer for improving light output efficiency.

FIG. 10 shows an eighth embodiment LED having a bottom electrode.

FIG. 11 is a view showing an LED having a ring-shaped electrode according to a ninth embodiment.

FIG. 12 is a tenth embodiment having a ring-shaped electrode and a bottom DBR.
It is a figure which shows LED of an Example.

FIG. 13 shows the LED mold structure of the eleventh embodiment
It is a figure shown with OF.

FIG. 14 is a graph showing calculations showing how coupling efficiency depends on lens distance and fiber distance.

FIG. 15 is a graph showing a calculation showing how coupling efficiency depends on lens distance and fiber distance.

FIG. 16 is a schematic view of a conventional visible LED.

FIG. 17 is a diagram showing a schematic near-field pattern of the LED of FIG. 16.

FIG. 18 is a schematic view showing a structure of a conventional LED.

[Explanation of symbols]

 1 substrate 2 n-GaAs buffer layer 3 n-clad layer 4 active layer 5 p-clad layer 6 block layer 7 current injection layer 8 high concentration p-cap layer 9 insulating layer 11 n-contact 12 light emitting surface 13 passivation 14 DBR 16 Bonding area 18 Block layer 19 GaAs thin film 21 Bonding area 22 Au plating 23 LED 24 Leaf frame 26 Lens 27 POF

Claims (13)

[Claims] 1. A first conductivity type buffer layer, a first conductivity type clad layer, an active layer, a second conductivity type clad layer, a second conductivity type thin film, and a second conductivity type current injection layer on a first conductivity type substrate. A surface-emitting diode in which a high-concentration second-conductivity-type cap layer is sequentially stacked, having a mesa structure in which the second-conductivity-type clad layer is etched, and leaving a central portion in the second-conductivity-type thin film. A surface emitting diode, wherein a block layer selectively oxidized to the side surface of the mesa is formed. 2. A first conductivity type buffer layer, a first conductivity type clad layer, an active layer, a second conductivity type clad layer, a second conductivity type thin film, a second conductivity type current injection layer, on a first conductivity type substrate. A high-concentration second-conductivity-type cap layer is sequentially stacked, and the second-conductivity-type clad layer is mesa-etched, and then the second-conductivity-type thin film is selectively oxidized to the side surface of the mesa leaving a central portion to form a block layer. A method for manufacturing a surface emitting diode, comprising depositing an insulating layer on the structure and providing a light emitting window on the top of the mesa. 3. A first conductivity type GaAs buffer layer and a first conductivity type (Al _{1 -X} Ga _X ) on a first conductivity type GaAs substrate.
_0.5 In _0.5 P clad layer, (Al _1-X Ga _X ) _0.5 I
n _0.5 P bulk or quantum well structure active layer, second conductivity type cladding layer, second conductivity type thin film Al _X Ga _1-X As
(0.9 ≦ X ≦ 1), second conductivity type Al _X Ga _1-X As
(0.7 ≦ X ≦ 1) A surface emitting diode in which a current injection layer and a high-concentration second conductivity type cap layer are sequentially stacked, having a mesa structure in which the second conductivity type cladding layer is etched. A surface emitting diode, wherein the second conductive type thin film is formed with a block layer selectively oxidized to a side surface of the mesa except for a central portion. 4. The surface emitting diode according to claim 1, wherein the diameter of the light emitting surface is 100 μm or less. 5. A first conductivity type buffer layer, a first conductivity type clad layer, an active layer, a second conductivity type clad layer, a second conductivity type current injection layer, and a high concentration second conductivity type on a first conductivity type substrate. 1. A surface-emitting diode in which a cap layer is sequentially stacked, wherein a high resistance region is formed in the second conductivity type current injection layer by ion implantation. 6. A first conductivity type buffer layer, a first conductivity type clad layer, an active layer, a second conductivity type clad layer, a second conductivity type current injection layer, and a high-concentration second conductivity type on a first conductivity type substrate. A surface emitting diode having a cap layer sequentially stacked, the surface emitting diode having a diameter slightly larger than a light emitting surface diameter and having a mesa structure etched up to the second conductivity type clad layer. 7. A first conductivity type buffer layer, a first conductivity type clad layer, an active layer, a second conductivity type clad layer, a second conductivity type current injection layer, and a high-concentration second conductivity type on a first conductivity type substrate. A surface-emitting diode in which a cap layer is sequentially stacked, wherein a ring-shaped first conductivity type or insulating block layer is formed in the vicinity of the second conductivity type clad layer of the second conductivity type current injection layer. A surface-emitting diode characterized by: 8. The surface-emitting diode according to claim 1, 5, 6, or 7, wherein the electrode on the light-emitting surface side has a ring electrode of one or more layers, and the ring is made of an insulating material transparent to emitted light. A surface emitting diode characterized in that a light emitting surface other than the electrodes is covered. 9. The surface emitting diode according to claim 8, wherein the insulating material is silicon oxide, silicon nitride, or alumina. 10. The surface emitting diode according to claim 1, wherein a distributed Bragg reflector (DBR) mirror is formed on the first conductivity type buffer layer. . Wherein said distributed Bragg reflector (DBR) mirror 12 or more pairs of _{AlAs / (Al 0. 5 Ga 0.5} ) 0.5
In _0.5 P or 20 pairs of AlAs / Al _0.5 Ga _0.5 A
s or 15 pairs of As _0.5 In _0.5 P / (As _0.5 Ga
_0.5 ) _0.5 In _0.5 P or 25 pairs of Al _0.5 In _0.5 P
/ Ga _0.5 In _0.5 P or 15 pairs of AlAs / GaAs
It consists of 1 or 5, 6, 7,
8. The surface-emitting diode described in 8 or 10. 12. A mold formed of a material having a refractive index of 1.6 or less, having a lens diameter of 0.6 mm or less, and a distance from a light emitting surface of a surface emitting diode to a vertex of a lens center of 0.75 mm or less. The surface emitting diode according to claim 1 or 5, 6, 7, 8, 10 having a lens. 13. A step index (SI) POF having a numerical aperture (NA) of 0.5 or less and a diameter of 1.0 mm or less and a distance between the lens and the lens is 7.5 mm or less in order to obtain a coupling efficiency of 80% or more. 13. The method according to claim 12, wherein
The surface emitting diode described.